In mid-December, Reda Ihtassine, research associate in our “Data Science” workgroup, was at the K-CAP 2025 conference in Dayton, Ohio (USA), to present his paper entitled "NL2LS: LLM-based Automatic Linking of Knowledge Graphs".
In this paper, he addresses the problem of automatically generating link specifications. The idea is that NL2LS automatically translates natural language into formal link specifications that are needed to link resources in knowledge graphs. This makes it much easier for non-technical users to create complex matching rules and automates a process that was previously highly dependent on experts. Co-authors of the paper are Asep Firmansyah, Nikti Srivastava, Ali Manzoor, Axel Ngonga and Mohamed Sherif.
The K-CAP 2025 conference took place from 10-12 December in Dayton, USA and is the thirteenth international conference on Knowledge Capture. K-CAP 2025 continues its tradition of bringing together an interdisciplinary community dedicated to advancing the science and technology of knowledge capture.
Read more about the conference here.
